Despite the significant progress of women in overall educational attainment during the last few decades in the United States and other developed countries, the proportion of females in STEM college majors and occupations is much smaller than the proportion of males and the proportion of individuals from low parental SES (socioeconomic status) backgrounds is much lower than those from higher SES backgrounds. In the proposed project, we will investigate whether one particular educational institution - single-sex high schools - help overcome gender and parental background equity issues in STEM learning by developing and maintaining interest in STEM among girls as they progress through school and job, and thus encouraging more gender diversity within the population of students interested in STEM careers. In addition to the major focus on girls, this project will also assess effects of all-boys schools on boys as well. A critical improvement of this project over previous literature is to use data from South Korea where both students and teachers are randomly assigned into single-sex vs. coeducational academic high schools in most of the country. The random assignment of students into high schools in Korea enables us to address selection biases from which most previous studies have suffered on the basis of observational data, and to draw causal inferences on the effects of single-sex high schools. Specifically, using four different data sets of nationally representative high school students and college graduates, we will address whether studying in all-girls high schools should increase 1) the performance of high school girls on standardized mathematics and science tests;2) the probability that high school girls expect that they will have STEM majors when they continue on to universities;3) the probability that girls subsequently actually have STEM majors when they continue on to universities;4) the probability that girls will enter STEM occupations when they complete their formal education. In addition to average effects of single-sex high schools, we will also address 5) whether these effects of studying in all-girls high schools depend on parental SES, for example, with stronger effects for girls from lower parental SES backgrounds. Finally, we will examine whether studying in all-boys'schools have parallel effects on boys. By examining both short-term and longer-tem effects of single-sex high schools on the basis of the nation-wide randomized data from Korea, the project thus promises significant improvements of our knowledge regarding whether single-sex high schools might be an important institution to address the paucity of females and low- SES students in STEM careers. Causal inferences on single-sex school effects drawn from our study can offer timely scientific evidence for recent educational efforts elsewhere. In 2006, for example, the US Department of Education established new regulations to make it easier for school districts to offer single-sex education. As a result, the number of public schools that are single-sex schools or offer single-sex classrooms within coeducational schools has soared from 3 in 1995 to more than 545 in September 2009. PUBLIC HEALTH RELEVANCE: The United States Department of Education has recently changed regulations to make it easier to establish single-sex public schools, based on the potentials of single-sex schooling for educating American children. Our project will inform public policy by drawing causal inferences on whether single-sex high schools promote participation of females and economically disadvantaged students in Science, Technology, Engineering and Mathematics (STEM) careers. If there is an increase in mathematics and science knowledge and focus on STEM majors by females and economically disadvantaged students, this surely will lead to increases in females and economically disadvantaged students working in areas related to health research and practice.